The relation between law and science is rather peculiar. Scientists on the whole have no interest in law whereas the law increasingly makes use of science. Both lay claim to being very good at discovering the truth though they do so in different ways. While the law makes use of evidence, it also relies on authority and precedence and morality, and it can frame and direct social policy. Science is not concerned with the good or justice, but just the way the world actually works. Science is the best way to understand how the world works; it provides value-free, objective, reliable knowledge. The final appeal in science is to Nature, and one's colleagues; in law it is to the House of Lords.

Law and science differ in at least one key area. Law is often concerned with the incidence of individual events and their causation. Science by contrast is rarely interested in single events but rather in general understanding, laws that apply to a wide variety of events.

The spinning of a coin can illustrate the differences that this involves. The scientist can predict, using classical mechanics, whether a coin will come down heads or tails, if the initial conditions are known, but these are rarely available. What the law might want to know is why it was a head rather than a tail. The analogy is with whether a chemical or treatment could have caused some damage or illness, or whether a child is likely to be abused. How does one deal with the probable risk of single events?

It is an issue of major importance as to how the courts should use expert scientific evidence and was discussed extensively at a rare meeting on science and law. What scientific evidence is admissible and how is it to be evaluated? A good criterion seems to be that the evidence should be helpful, but here we meet an apparent overconfidence of judges. For it is argued that when it comes to psychological issues like child rearing or marital conflicts where judge and jury have personal experience, then common sense should be the guide and science is not relevant. But, as I have so often argued, most science goes against common sense.

The Americans have put down some useful guidelines for what science should be admissible in order to exclude junk science from the courts. These criteria include testability, satisfactory peer review, some indication of the possible errors, and how widely the science is accepted by the scientific community. Even so, judges have some difficulty and on one occasion when the scientific evidence for chemical pollution of a river was overwhelming, the judge still wanted to know how many dead fish had been found. There were no fish left to die.

Particular problems arise in the social sciences because the phenomena are so complex. For example rape trauma syndrome as a medical condition was promoted by the women's movement, but whether it is a medical condition is hard to establish. And if a woman does not exhibit the standard symptoms does this imply that her claim to have been raped is false? There is also the feeling that junk science may enter the court room when religious movements are accused of brainwashing.

In our adversarial system, experts are inevitably biased towards one side. Money was rarely mentioned at the meeting, but I cannot help speculating that if expert witnesses were only paid expenses, they would be less inclined to offer their services, and much junk science and dispute would disappear from the courts.